Advancing N₂O flux chamber measurement techniques in nutrient-poor ecosystems

Abstract. Nitrous oxide (N₂O) is the third most important greenhouse gas, with its atmospheric concentration rising from 273 parts per billion (ppb) to 336 ppb since 1800, primarily due to agricultural activities. However, nutrient-poor natural soils, including those in the sub-Arctic, also emit and consume N₂O. These soils have not been studied extensively, partly due to challenges in reliably detecting low fluxes. Methodological limitations were largely influenced by the available instrumentation; the lack of portable gas analysers for N₂O with adequate accuracy led researchers to rely on manual air sampling from closed flux chambers, followed by laboratory analysis using gas chromatographs (GC). In this study, we utilized a fast-responding portable gas analyser (PGA; Aeris N₂O/CO₂) combined with a custom manual chamber system, which includes both transparent (light) and opaque (dark) measurements, to effectively measure low N₂O fluxes from a nutrient-poor sub-Arctic peatland. We assessed the analyser's performance under low-flux conditions, evaluated the effects of chamber closure times, and compared linear and non-linear models for quantifying concentration gradients. Additionally, we analyzed flux rates based on high-frequency in situ observations against a method that randomly selects discrete samples from the full time series, simulating a GC-based approach. Our results indicate that the PGA can reliably detect and compute low N₂O flux rates, averaging 12.9 ± 28.4 nmol m⁻² h⁻¹ under light conditions and -46.1 ± 38.2 nmol m⁻² h⁻¹ under dark conditions, depending on chamber closure time. The majority of fluxes (88 % for light and 74 % for dark measurements) exceeded the minimum detectable flux (MDF), which was 14.5 ± 1.05 nmol m⁻² h⁻¹ for light and 14.7 ± 1.08 nmol m⁻² h⁻¹ for dark measurements. Our comparison of chamber closure times (3–10 minutes) showed that a 3-minute closure may be inadequate for capturing low N₂O fluxes during light measurements, while closure times of 4–10 minutes yield more reliable results. For dark measurements, where N₂O uptake peaked with shorter closure times, we recommend a closure time of 3–5 minutes unless data are limited; in such cases, longer times may help capture fluxes above the MDF. In our study, all N₂O fluxes were calculated using the non-linear model or corresponded with the linear model when data exhibited a linear distribution. Compared to PGA-based flux calculations, GC simulations underestimated N₂O fluxes when using 3–6 samples. Therefore, we conclude that fast-responding analysers may be more suitable for measuring low N₂O fluxes, enhancing our understanding of the complex dynamics of N₂O emissions.